[Debugging] Sola Sound - Tone Bender Mark I

ivan H · Post by **ivan H** » 08 Jan 2016, 01:56

Hi, I've been a member here for a couple of years but this is my first post (I think). I recently built a mk1 (sola sounds, not the Hurst version) using OC75's so thought I'd revive this thread with a little info on how I went about the transistor selection/biasing process in the hope others may benefit from it in future. This was my first mk1 build so I learned a lot from this forum (thanks), but I found excerpts from Meastro FZ-1 patent elsewhere that helped immensely, so here goes.
I built up a test circuit on Vero board using Sinner's layout, but spaced it out so that I could seperate each stage & thus take the signal out after Q1 & 2, or inject it back in front of Q2 & 3 (sockets). I started using the stock resistance values (180k from -9v to q2 base). So, choosing Q1. As recommended earlier in this thread I looked for a transistor that gave a clean reproduction of the input signal. The FZ-1 patent says this stage doesn't amplify but as the mk1 supposedly came about from a request for more sustain out of an FZ-1, I chose one that did boost the signal a little & had good tone. This is important as I noticed the tone varied between devices somewhat. Ok, Q2. From the FZ-1 patent this is the stage that amplifies the signal. The transistor should be biased just a little above cutoff. The 1k8 limiting resistor (2k2 in Hurst variant) between the transistor's base & the "attack" pot is what prevents cutoff when the pot is at minimum. So selecting this transistor is pretty easy. I first paralled the 1k8 resistor with another 1k8, then auditioned transistors to find one that gave the faintest signal out with the pot at minimum & good amplification as it was turned up. It will distort etc as it's turned up but that's ok. I then removed one 1k8 resistor & checked it again. All good. Ok, Q3. This is the real "fuzz" producing stage. It should be biased on the right on the threshold of cutoff. Applied signal then drives it into cunducting. To select this transistor I first set the trimmer for Q3 to 8k2 & the attack pot to minimum & started auditioning transitors, first just looking for one that would "work" of sorts. Now if you are lucky you'll find one that will work straight off, be quiet & allow use of the full rotation of the attack pot. Most likely you will have some noise being produced with the attack at minimum though it'll fuzz ok, but as the attack is turned up (it'll fuzz more) it will reach a point where the noise gets too much. What you need to do is lower the trimmer resistance (no applied signal) until it just quietens the noise being produced. I got lucky & found a transistor that would at least work with the stock 8k2 & allow full attack range, tho it was noisey. With 7k5 it was better, 6k8 its great. So with it all working properly you can tweak values to get your prefered fuzz tone from it. Places to tweak have been mentioned earlier in this tread. The 8k2 on Q1, the 180k (470k Hurst variant) from -9v to base on Q2. Increase to lower distortion. I found 240k to be best, which I believe was sometimes used in originals. The 1k8 resistance between Q2's base & the attack pot can be tweaked for the minimum fuzz setting. Hope this is helpful for future builders. Cheers

ljn · Post by **ljn** » 08 Jan 2016, 10:26

I'm having trouble with my Mk. I build. It's working, but it has too much low end and a short, fatty decay. I used some AC128 transistors I ordered from eBay. For Q1, I have hfe 60 and leakage of 0.10 ma. Q2 has a hfe of 73 with 0.06 ma of leakage and Q3 has a hfe of 62 with 0.06 ma of leakage. These weren't a matched set. I just bought a bag of transistors because they were cheap. There were 24 AC128 and one AC176. I tried most of them with no noticeable difference. I'm thinking of ordering a matched set of either OC75, or AC125. There's also OC42, but I've never heard of those before.

ivan H · Post by **ivan H** » 09 Jan 2016, 09:32

Hi, I had ordered some AC125's that turned up yesterday afternoon. As I had to lower the 8k2 resistor on Q3 a little I thought I'd try a few of the AC125's. I ended up changing both Q2 & Q3 to AC125. Now with all stock values, still with 240k from -9v to Q2's base & no resistor on the attack pot its working very well & I can use the full rotation of the attack pot, so I'll give you some numbers. Q1, OC75, hfe 70, leakage 210. Q2, AC125, hfe104.8, leakage 328. Q3, AC125, hfe 66.5, leakage 171. Tomorrow I'll measure the voltages at various "attack" pot positions & post for you. I must agree with other that have said that it really comes down to finding the right transistors, especially for Q2 & Q3. Q1 isn't so critical, so long as it gives you a clean reproduction of the input signal & has nice tone. It doesn't seem to interact so much with the others, but Q2 & 3 definitely so. Hope you get it sorted, I'll get back with those voltages. Cheers

ivan H · Post by **ivan H** » 09 Jan 2016, 09:54

Oh, while I think of it, the OC75's that I had had as Q2 & Q3 were; Q2, hfe100, leakage 289. Q3, hfe 74, leakage 250. This had worked with the 8k2 but was noisey & required the 8k2 on Q3 reduced to 6k8 to get decent performance. Its definitely not a bassy fuzz. I'd suggest listening to the signal coming out of the first stage (straight from the + lead of the 25uf electrlytic into your amp) to make sure you are getting a nice clean tone from it, first. Then once that's sorted you can concentrate on the next 2 stages. Hope this helps. Cheers

ivan H · Post by **ivan H** » 09 Jan 2016, 11:31

Ok, so I checked the voltages for you, so here they are. I stuck a new battery first. It read 9.98v straight out of the wrapper.
Q1; C, -9.89v. B, -1.92v. E, -1.93v. These stay the same no matter what position the attack pot is.
Attack at minimum;
Q2; C -9.59v, B, -0.08v.
Q3; C, -9.30, B, -0.03

Attack at 9 O'clock
Q2; C, -4.22v, B, -0.16v.
Q3; C, -9.16v, B, 0.03v.

Attack at 12 O'clock
Q2; C -2.30v. B, -0.17v.
Q3; C -9.11v. B, -0.03v.

Attack at 3 O'clock
Q2; C-1.92v. B, -0.18v
Q3; C -9.08v. B -0.03v.

Attack at maximum
Q2; C -1.83v. B -0.18v
Q3; C -8.94v. B -0.03v

These readings are with a 240k resistance between the -9v rail & Q2's base. They vary depending on this resistance value. Should you have 180k there, Q2's collector may come down a little more as the attack is turned up. If you have 470k there (as in the Hurst variants), Q2's collector should only drop to a bit below -5v with the attack at maximum. I believe that 240k was a value used in some original mk1's. Hope this helps. Cheers

ljn · Post by **ljn** » 09 Jan 2016, 12:28

Yes, this will help. Thank you very much. I definitely need some transistors with higher gain and leakage then. I've been looking at some matched sets on ebay, and I was really wanting to go with the OC75 set, but I would settle for AC125. Those are cheaper, it seems. You mentioned a 240k resistor. I've never seen that particular value. Could someone use a 220 k instead with similar results. I'm trying to keep mine as close to the original values as possible. Of course, the part types are impossible for me to match. I'm only concerned with the values of the caps and resistors staying close to stock values. I used a 22 uf electrolytic cap in mine because 25 if is difficult to find. The only ones around here are in one of my amplifiers. Thanks again. I'm gonna take a break from this circuit until I can get some better transistors.

ljn · Post by **ljn** » 09 Jan 2016, 12:30

I forgot to mention that I do have a 180k resistor between -9v and Q2 base.

Electric Warrior · Post by **Electric Warrior** » 09 Jan 2016, 18:37

ljn wrote:You mentioned a 240k resistor. I've never seen that particular value. Could someone use a 220 k instead with similar results. I'm trying to keep mine as close to the original values as possible.

Sure.

They tweaked that resistor a lot when building vintage units. There is evidence of resistors being cut out of the circuit and replaced with different values. Just because one particular unit had a 180k resistor doesn't mean that this value is gonna work for you.
There aren't that many authentic vintage units around, but so far seems the range of values they used is from 180k to 560k.

I'm only concerned with the values of the caps and resistors staying close to stock values. I used a 22 uf electrolytic cap in mine because 25 if is difficult to find.

No big deal. Electrolytics had and still have wide tolerances.

ivan H wrote: These readings are with a 240k resistance between the -9v rail & Q2's base. They vary depending on this resistance value. Should you have 180k there, Q2's collector may come down a little more as the attack is turned up. If you have 470k there (as in the Hurst variants), Q2's collector should only drop to a bit below -5v with the attack at maximum. I believe that 240k was a value used in some original mk1's. Hope this helps. Cheers

I've seen a 270k in one.

I built my clone using a 470k. I had tried 1k8 at first, but it had fizzy decay. With the 470k that wasn't much of a problem anymore and I was able get a good range on the attack knob. Measures around -6V on Q2's collector with the knob turned up.

ljn · Post by **ljn** » 09 Jan 2016, 18:54

Okay. Good to know. I do have another question about the typical behavior of this circuit. Does it clean up any when the guitar volume is backed off? The only examples of this effect I've really heard are on some Yardbirds tunes, so I have very little information on how this circuit is supposed to act. I have seen some demos on YouTube, but couldn't really tell much about the pedals. It seems like every demo I watched had the pedal going through a tube amp. That doesn't really help solid state amp users, such as myself. I'm using an epiphone Les Paul standard with an early 70s Kasino U100-P amp. Thanks again for the info about the voltages and the resistor values. Much appreciated.

ivan H · Post by **ivan H** » 10 Jan 2016, 00:54

Hi, yes it will clean up some by turning your guitar volume down & the response also changes. I just gave it a quick try for you. 2012 Gibson '60's tribute SG (bridge p'up, A5, 11.6k) into a '68 spec 50 watt plexi clone. Fuzz level unity (a little below 12 O'clock), Attack at minimum, guitar volume just above the threshold where it drops right away to nothing. Yes, it cleans up, but you can tell its thru an effect. Single notes are better. If played with a bit of attack it gets a bit fuzzy on attack & cleans up. Not much in the way of sustain. Bar chords have a bit of fuzzy attack. But this is with the guitar unrealistically low. Turning it up as its turned up the fuzz becomes more & more apparent, but yes, it does clean up. Again, this is with the attack at minimum. I've got things to do at the moment so will try at different levels of attack later. I cant help with how it goes into an SS amp. Cheers

ljn · Post by **ljn** » 10 Jan 2016, 01:47

Thanks for taking the time to find out. If it cleans up through your tube amp, then I'd say it should also clean up throat solid state amp. Mine doesn't clean up, but I guess that's a side effect of being misbiased. Even with the attack at minimum, mine has been giving a full on fuzz sound. As the attack is turned up, the signal gets nastier and gives a short, crackling decay. I don't think that's normal. Thanks again.

ivan H · Post by **ivan H** » 10 Jan 2016, 02:01

No problem. The way I use mine (even though I haven't had it long) is to set the attack where it sounds good & the level so it pushing the amp (boosting). I then leave it & use the guitar's volume to control it. Don't know if that's how its meant to be used but it works for me. For OC75's (& other types too), try Steve Daniels at Smallbear electronics (online store). I've had good results buying from him & he's good to deal with. Cheers

ljn · Post by **ljn** » 10 Jan 2016, 04:17

I was thinking of ordering a set from LIC pedals. Even after the circuit is straightened out, I'll still have to order some pots and a bypass switch for it. I'm also planning on building an authentic style enclosure to put it in. There's a place where I can get sheet metal locally. I'm even going to paint it in the original colors with some Rust-Oleum Hammered spray paint. Love that stuff. It has a long way to go. Thanks again.

ljn · Post by **ljn** » 10 Jan 2016, 06:29

Okay. I took the 180k resistor out and replaced it with a 270k. It improved the performance a great deal, but the overall tone is still very dark. Could the resistor on Q1 emitter be responsible? It's a vintage part that I pulled from an old tube radio chassis and it did read kind of high. Around 11k, if I remember correctly. I think I'll try replacing it and see what happens, but I still think my transistors are the problem.

ivan H · Post by **ivan H** » 10 Jan 2016, 07:20

It would have some affect on tone, that resistor is biasing Q1. Any affect on tone there is going to carry through the whole chain. I've boxed mine up in a 125b so it will fit easily on my board, otherwise an original style enclosure would have been nice. I did it on perf board, point to point with a ground bus. I used tropical fish (NOS) caps & an axial 25uf spague cap that I had. I used 1% tolerance metal film resistors. I don't like carbon comp in pedals, especially with germanium's. Aside from drifting up with age they are notoriously noisey. I use them sometimes in amp builds, but only in a few places. I find they need a large voltage swing across them to get the "mojo" affect. In a pedal like this using them, especially used ones is asking for noise & mis- bias problems, but that's only my opinion. Try a resistor that measures close to 8k2. Cheers

ivan H · Post by **ivan H** » 10 Jan 2016, 07:27

Sorry, I just re read your post, yes, I also noticed that each transistor sounded a little different when I was looking for Q1. Some had good tone, others not so good, even though not distorting. Cheers

ljn · Post by **ljn** » 10 Jan 2016, 08:44

I replaced the 8k2 resistor with another that read exactly 8.2k ohms. It had no affect on the tone. The only questionable one left in the circuit is the 1meg on Q1's base (another vintage part I've had for several years) . The only other carbon comps in there are the 2k2 on Q2 collector (which was tested and read very close to spec ) and the 1k8 that goes to the attack control ( wasn't tested) . There's not much else that could be going wrong here. I'll try replacing the 1 meg and the 1k8 and see if that helps. Thanks again.

ljn · Post by **ljn** » 10 Jan 2016, 09:31

I just swapped out those two other resistors and as before, it had no effect. Except there was a very subtle increase in the midrange response that I didn't notice at first. By the time I get the right transistors, this one should be ready for them. That's the only thing that's left. So which is better? OC75 or AC125? At this point, I'm only concerned with getting it to sound good. The reason I ask which transistors are better is because I've read that certain germanium are known for a particular tonal quality. I should've thought of that before buying AC128. It's not a total loss, though. I'm definitely gaining experience. Until recently, I had only been working with silicon. Oh, my input cap should probably be mentioned, as well. Yes, it's another vintage part from an old radio. A big grey axial lead rated for 800 volts. Do you think that could be causing an issue with the tone? Thanks again.

ljn · Post by **ljn** » 10 Jan 2016, 10:23

I found out what was causing the muddy sound, and it's a little embarrassing. I had the electrolytic cap in backwards. Once I noticed it's orientation, I removed it and put it back in the correct way. The tone is now alot brighter and it actually sounds similar to what I've heard on those Yardbirds songs. It could still do with some higher gain transistors, but I'm digging the sound.

ivan H · Post by **ivan H** » 10 Jan 2016, 10:43

Well, capacitors can color the tone in their own way, which is why people often prefer different types in different positions, but they don't usually go bad like old electrolytics do, though they can go leaky, which shouldn't really be a problem with an 800v cap in this application. I think at least an OC75 for Q1, as that's what they used. For the other two I'd be happy with either OC75 or AC125. Whatever you can get easily. Smallbear sells a set of OC75's for a mk1. As I said earlier, I've had good results buying from him. I'm down under in Oz & he gets orders here quick as anyone I order through. You may need to tweak a little but not too much. Don't give up, I was auditioning transistors on & off for nearly two weeks (& taking notes on each combination) before I got it to where I was happy. Its worth it in the end, I think it'll knock the mk11 off my board. I've just started on a Selmer Buzztone. Cheers